A Comparison of Modality Tests Based on Real Life Data Applications | Proceedings of the Pakistan Academy of Sciences: A. Physical and Computational Sciences 2019

Author(s)

Rana Muhammad Imran Arshad Department of Statistics Govt., S.E. College,Bahawalpur,Pakistan

Shahina Imran Department of Economics Govt., S. E. College,Bahawalpur,Pakistan

Farrukh Jamal Department of Statistics Govt., S.A. Postgraduate College,Dera Nawab Sahib, Bahawalpur,Pakistan

Gamze Ozel Department of Statistics, Hacettepe University,Ankara,Turkey

Abstract

Nonparametric tests for the modality are known as distribution-free tests to evaluate the evidence about homogeneity in a population. Silverman's bandwidth test by Silverman [1], the Hartigan's dip test by Hartigan [6], Hartigan and Hartigan [7], Proportional Mass (PM) test by Cavallo and Ringobon [4] and Excess Mass (EM) test by Muller and Sawitzki [9] are very popular modality tests in literature. The researchers of these tests claimed that their tests perform well to detection of unimodality, bimodality and multimodality. This study focusses on the comparison of these tests based on real data applications. The results show that Silverman's bandwidth test is the best test for detecting modality whether it is unimodality, bimodality or multimodality.

Publication Details

Page(s) 69-75

DOI DOI not available

Published Journal: Proceedings of the Pakistan Academy of Sciences: A. Physical and Computational Sciences, Volume: 56, Issue: 1, Year: 2019

Keywords

Excess Mass test unimodality Nonparametric tests Hartigans dip test Proportional Mass PM test Silvermans bandwidth test Real life data application

References

References are not available for this document.

Citations

Citations are not available for this document.